The AI Pilot Graveyard Has a Name: Why AI Projects Fail on Leadership — Not on Technology

Fresh from IFA and reported in the FAZ: Germany's digital minister Karsten Wildberger is calling for more courage and speed on AI — Germany, he argues, can't afford to get lost in hesitation while global competitors race ahead. He's right: endless second-guessing gets nobody anywhere. But speed alone isn't a strategy. Slam the accelerator without steering, and you just reach the pilot graveyard faster.

What companies urgently need right now are clear ground rules — internal ones. This isn't about more external regulation or paralyzing bureaucracy. It's about practical, workable governance: orientation, guardrails, and trust for the workforce. External regulation sets limits — leadership and internal rules create the ability to act.

The numbers show just how far wish and reality currently diverge. Germany's KfW SME Panel puts it plainly: 20 percent of companies now use AI — up from just four percent a few years ago. But look closer, and the picture thins out: on average, that's a meager two applications per company, and fewer than a quarter have any fixed rules for them. Lots of starting, not much depth. McKinsey records rising scaling of AI initiatives for summer 2026 — from 38 to 44 percent of companies surveyed — but only 37 percent of those report a measurable contribution to EBIT, and just 6 percent qualify as "AI high performers." Germany's Bitkom digital association adds, for the Mittelstand specifically: 41 percent already use AI, a third of them at higher cost than expected, and one in five already reporting job cuts as a result.

These studies, entirely independent of one another, arrive at the same pattern: more use, barely more return. The obvious reflex is to blame the technology — the wrong tool, a missing data model, not enough compute. That reflex is exactly what leads into the dead end. The World Economic Forum describes five distinct company readiness types and explicitly recommends cultural change and transparency ahead of a technology focus. Fitting that picture, recent Glassdoor data show positive AI sentiment collapsing from 81 to 43 percent — and a Notion study shows a gap between leadership and staff: 60 percent of decision-makers consider their company AI-ready, but only 36 percent of employees agree.

From my own years in leadership, I know this pattern well: technology can be bought, acceptance can't. When leadership and staff give different answers to the same question — are we AI-ready? — that's not a communication problem another training session will fix. It's a leadership and governance problem: who decides what happens with AI? Who's accountable when a pilot fails? And who actually brings the workforce along, instead of merely informing them? These are exactly the questions the AI Governance Discovery answers — not with another tool ranking, but with named responsibilities and a process that involves leadership and staff together.

So the pilot graveyard does have a name — and it isn't AI. Where your company stands on this leadership question is what the free AI Governance Check shows, in five questions.